Techniques for evaluating a video rate selection algorithm over a completed streaming session

Abstract

In various embodiments, a hindsight application computes a total download size for a sequence of encoded chunks associated with a media title for evaluation of at least one aspect of a video streaming service. The hindsight application computes a feasible download end time associated with a source chunk of the media title based on a network throughput trace and a subsequent feasible download end time associated with a subsequent source chunk of the media title. The hindsight application then selects an encoded chunk associated with the source chunk based on the network throughput trace, the feasible download end time, and a preceding download end time associated with a preceding source chunk of the media title. Subsequently, the hindsight application computes the total download size based on the number of encoded bits included in the first encoded chunk. The total download size correlates to an upper bound on visual quality.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS[0001]This application claims the benefit of United States Provisional patent application titled “ADAPTIVE BITRATE EVALUATION,” filed on May 24, 2018 and having Ser. No. 62 / 676,246. The subject matter of this related application is hereby incorporated herein by reference.BACKGROUNDField of the Various Embodiments[0002]Embodiments of the present invention relate generally to streaming video technology and, more specifically, to techniques for evaluating a video rate selection algorithm over a completed streaming session.Description of the Related Art[0003]A typical video streaming service provides access to a library of media titles that can be viewed on a range of different client devices, where each client device usually connects to the video streaming service under different connection and network conditions. In many implementations, a client device that connects to a video streaming service executes an endpoint application. The endpoint appl...

AI Technical Summary

Benefits of technology

The patent text describes a new technique that can more efficiently determine the quality of visual experience for a media title during playback. This is done by correlating the quality to the total size of the media title, which is made up of smaller chunks of data. This approach overcomes difficulties in previous methods that were time-consuming and computationally demanding. The disclosed technique reduces the amount of time and resources needed to evaluate the quality of video across different network environments and device capabilities.

Problems solved by technology

One drawback of this first approach is that optimal video rate selection is an NP-hard problem, where the abbreviation “NP” stands for non-deterministic polynomial time.

As is well-understood, an NP-hard problem cannot be solved efficiently using known techniques.

Because executing the optimal video rate selection algorithm for each completed streaming session is highly inefficient in view of the NP-hard problems, computing the numerous optimal QoE metric values required to comprehensively evaluate the video rate selection algorithm is usually prohibitively time consuming.

Given the above drawbacks, many video streaming service providers do not attempt to compare the performance of a video rate selection algorithm to the performance of an optimal video rate selection algorithm.

One drawback of this second approach is that a current video rate selection algorithm in use as part of a video streaming service may perform reasonably well in many operating scenarios.

Consequently, developing a candidate video rate selection algorithm that can outperform the current video rate selection algorithm can be a challenging, time-consuming, and primarily manual process.

Further, if the candidate video rate selection algorithm fails to outperform the current video rate selection algorithm, then the process of comparing the candidate video selection algorithm to the current video rate selection algorithm provides no guidance on how to generate a new candidate video rate selection algorithm that actually outperforms the current video rate selection algorithm.

Images